- Ancient Writing Systems
- Further information
The language (or languages?) of the Libyco-Berber texts remains inadequately understood; not only is the corpus too small and too stereotyped to provide a full picture of the language, but the wide geographical spread of the texts implies the possibility of significant variation. However, a number of bilingual texts make it possible to determine some of its characteristics with relative certainty, and point to a close relationship to the modern Berber languages. The longest of these bilinguals come from the Numidian city of Dougga (tbgg) in modern-day Tunisia, and presumably reflect the dominant language of the Numidian kingdom.
Nouns may form their plurals with a suffix -n (modern Berber m.pl. -ən), eg nbbn "workers (woodcutters?)".
In genitive constructions, the possessed precedes the possessor. Family terms are followed directly by the genitive (eg Znn w Yrnbt "Zanan son of Yarnabat"), whereas inanimate genitives are marked with a particle n as in modern Berber (eg nbbn n šqrh "workers (cutters?) of wood".) The third person singular possessive is marked by a suffix -s (cp. Kabyle -is.)
In what may be appositive constructions, where a noun indicating profession precedes a personal name, the profession has -h suffixed; where it follows the name, it has -t1 suffixed. The latter has been tentatively compared to the stative ending -ət of some modern Berber languages.
d "and" links noun phrases, as generally in modern Berber, and possibly (in RIL 3) clauses.
Few verbs have been unambiguously identified in the texts. Comparison with modern Berber languages suggests that ṣkn "built" in RIL 2 is to be analysed as ṣk "build" plus -n, marking 3pl subject agreement.
The words whose meaning is known with something approaching certainty, mainly through bilingual inscriptions, are few; the list here is nearly exhaustive.
The formula X w-Y "X son of Y", still used in some Berber languages, identifies people both in official inscriptions and on gravestones - many gravestones include little besides such a name, occasionally expanded to X w-Y Z "X son of Y of the tribe Z" (eg RIL 451.) ). .) A few gravestones show a different word between the two personal names, plausibly interpreted as a kinship term based on Berber comparisons: wlt "daughter (of)" (modern Berber wəlt), and, more rarely, mt "mother (of)" (modern Tuareg ma).
A number of political offices are mentioned in the Dougga inscriptions; including gld "king" (modern Berber a-gəllid), mwsnh (translated as "chief of the hundred"), gldmṣk (translated as "chief of the fifty"), and the Punic loanword šfṭ "suffete". Occasionally more prosaic professions are mentioned, such as nbb "worker (woodcutter?)" (RIL 1). Materials include šqrh "wood" (in which the phonetic value of "q" is guesswork, but cp. modern Berber asɣar) and zlh "iron" (modern Berber uzzal).
The Punic translation of RIL 2, combined with comparison to modern Berber, allows sbs- to be plausibly identified as "year". If this is correct, then proto-Berber *ww corresponded to b in Libyco-Berber, as in modern Zenaga in Mauritania, but unlike most other surviving Berber languages, where it has typically become *gg or ggʷ.